The overall goal of this research program is to develop spectroscopic- based glucose sensors that can be used in the treatment and control of diabetes. this sensing approach will be based on near-infrared (NIR) spectroscopy, state-of-the-art optical fiber technology, and advanced computer-based data analysis. The focus of the proposed work is the design, construction, and evaluation of a noninvasive sensor that will allow direct measurement of in situ blood glucose levels without collecting a blood sample. The proposed research is built upon a series of experiments designed to show conclusively that NIR spectroscopy, coupled with computer-based data analysis, can be used to determine glucose quantitatively in the 1 - 20 mM concentration range in biological matrices. The work described in this grant application consists of four major components: (1) extension of our initial feasibility studies to increase the complexity of the sample matrix; (2) investigation of the impact of instrumental parameters on the proposed analysis; (3) development of optimized data analysis algorithms for extracting glucose information from the collected NIR spectra; and (4) development of glucose sensing probes based on optical fiber technology. Throughout the proposed work, the principal focus of our research efforts will be to develop precise and accurate glucose sensing methods that are amenable to practical implementation.